Search bioRxiv⌕ Search

Biology subjects

Galvez-Merchan, A.

Publications and source records attributed to Galvez-Merchan, A..

5 recordsLinked to original sources
bioRxiv2024

Algorithms for a Commons Cell Atlas

Cell atlas projects curate representative datasets, cell types, and marker genes for tissues across an organism. Despite their ubiquity, atlas projects rely on duplicated and manual effort to curate marker genes and annotate cell types. The size of atlases coupled with a lack of data-compatible tools make reprocessing and analysis of their data near-impossible. To overcome these challenges, we present a collection of data, algorithms, and tools to automate cataloging and analyzing cell types across tissues in an organism, and demonstrate its utility in building a human atlas.

bioinformatics↗
bioRxiv2024

A human commons cell atlas reveals cell type specificity for OAS1 isoforms

We describe an open source Human Commons Cell Atlas comprising 2.9 million cells across 27 tissues that can be easily updated and that is structured to facilitate custom analyses. To showcase the flexibility of the atlas, we demonstrate that it can be used to study isoforms of genes at cell resolution. In particular, we study cell type specificity of isoforms of OAS1, which has been shown to offer SARS-CoV-2 protection in certain individuals that display higher expression of the p46 isoform. Using our commons cell atlas we localize the OAS1 p44b isoform to the testis, and find that it is specific to round and elongating spermatids. By virtue of enabling customized analyses via a modular and dynamic atlas structure, the commons cell atlas should be useful for exploratory analyses that are intractable within the rigid framework of current gene-centric cell atlases.

bioinformatics↗
bioRxiv2023

Quantitative assessment of single-cell RNA-seq clustering with CONCORDEX

The rapid advancement of spatially resolved transcriptomics (SRT) technologies has facilitated exploration of how gene expression varies across tissues. However, identifying spatially variable genes remains challenging due to confounding variation introduced by the spatial distribution of cell types. We introduce a new approach to identifying spatial domains that are homogeneous with respect to cell-type composition that facilitates the decomposition of gene expression patterns by cell-type and spatial variation. Our method, called concordex, is efficient and effective across technological platforms and tissue types, and using several biological datasets we show that it can be used to identify genes with subtle variation patterns that are missed when considering only cell-type variation, or spatial variation, alone. The con-cordex tool is freely available at https://github.com/pachterlab/concordexR.

bioinformatics↗
bioRxiv2022

Depth normalization for single-cell genomics count data

Genomics data analysis requires normalization of feature counts that stabilizes technical variance, accounts for variable cell sequencing depth, and preserves monotonicity of within-cell feature abundances. We show that normalization via an optimal variance stabilizing transform for negative binomial count data followed by a proportional fitting step (PFlog) is the only feature-relabeling-equivariant method satisfying the three desiderata. We demonstrate superior performance of this method, which is equivalent to a shifted centered-log ratio transform, in comparison to other normalizations on numerous benchmarks across hundreds of single-cell RNA-seq datasets. We further show that both the shifted-log scale and centered-log ratio geometry are important for preserving PCA and k-NN structure.

bioinformatics↗